Liquid in a cyclone

[Svetlin.Nikolov]

I will try to pick it and will let you know. I think it's possible to make it work with Phoenix 4.

Cheers!

[JuhaHo]

Thanks

I had yesterday meeting with the client. They told me to make big changes on simulation results - almost nothing is correct in the current simulation. I think I'm back at square one. Any ideas for totally new way to solve this? Is there a way to improve Phoenix fluid dynamic results? Anyway, my current results are the ones, that don't explode whole system. Change a tiny bit and Phoenix will explode all over the place.

I don't know what simulator they are using, but their simulations are much more physically correct.

On the screenshot below You can see, that the liquid starts going up at wrong place. In real world it goes all the way down and then most of it starts going up in the middle, while small amount of it goes to pipe down and flushes the particles there. The particles should stay inside the liquid and stop acting like a string of pearls outside of it. The movement of liquid is way too slow.

Any ideas how to fix these issues?

[JuhaHo]

Actually, it seems I'm after all running Phoenix 5. On license manager there's Phoenix universal. I didn't realize it means R5. So I'll download the latest nightly to see if it makes any difference. Anyway I have long way to go to get acceptable results with my simulation.

[Svetlin.Nikolov]

Hey, we just released Phoenix 5.10, so it would be best to install that: https://download.chaos.com/?platform=49&product=46

[Svetlin.Nikolov]

if the simulation explodes, this is absolutely a problem and is not normal. Would it be possible to send over the scene?

[JuhaHo]

Hey, we just released Phoenix 5.10, so it would be best to install that: https://download.chaos.com/?platform=49&product=46

Thanks,

I just installed it and waiting to see first results tomorrow.

Also, I have one question. As we know when rendering moving liquid, it's impossible to see it moving since. I was thinking if it's possible to render the same liquid simulation with smoke settings. I think it's easier to see how it's moving. If so, what would be the best practice for that?

[JuhaHo]

I just want to make sure, if the multipath option is available on the update, or do I need to use nightly version.

[georgi.zhekov]

Hey,

The multi spline support is already in the latest update of Phoenix 5.10. No need to use a nightly, just get the latest build from the chaos.com download section.

As for rendering liquid as smoke - you can pick the liquid particles in a Particle shader and render them as Fog - but this won't make the simulation behave like smoke though - it will just be rendered as volume - so it's best to try if this works for your scene.
The other option would be to make an additional smoke simulation and then render it together with the liquid.

[JuhaHo]

Thanks,

Regarding smoke, I was thinking it would make the liquid movement more visible. At the moment, liquid voxels are moving quite fast in the system, but in rendered animation they seem to be tumbling on spot. What if I add a copy and render it as smoke inside the liquid?

At my latest simulation the things are going a bit worse on particle point of view. Is there any ideas on how I can have a better distribution inside the fluid.

[georgi.zhekov]

You mean better distribution of the highlighted particles? How are they born - at some point you were using tyflow, is this still the case?

As for the smoke - sure, give it a go and check how it looks.

[JuhaHo]

You mean better distribution of the highlighted particles? How are they born - at some point you were using tyflow, is this still the case?

As for the smoke - sure, give it a go and check how it looks.

Thanks,

Particles are born at the same place where the liquid emitter is. They should be following the liquid to the bottom of the cyclone where the liquid starts rising up in the middle, but the particles being heavy should fall to the particle trap under the cyclone.

[georgi.zhekov]

Can you show us a video of how the simulation behaves? Also is it possible to share the scene - this way we will be able to suggest the best approach for your setup?

[JuhaHo]

I think I cannot share any renderings or models. I just have to try different settings and let Phoenix simulate a week or so until I see how beginning to look like. If it's better, I transfer it to the render scene. Then after week or so rendering, I can show it to client, who probably send me back to square one.

I'm just looking a way to get a bit of realism into my simulation, so it could be accepted by the client side. Now I'm nowhere near to that.

I think, if there were a way to simulate pressure changes, would help a lot.

[JuhaHo]

I modified the geometry at the bottom to reduce the counter pressure. It looks a bit better now. I hope the particles continue their travel down when the liquid starts traveling up. There is a conical collision shape in the middle for particles only. It has been excluded from Phoenix simulation. I'll let the simulation run like this for a while to see where this leads to.

[JuhaHo]

Still struggling with the scene. I wonder if the latest update broke something, because simulation eats up all the memory. I'm running this on my secondary workstation with only 32 GB of RAM. I think I have relatively coarse resolution. Is there anything I should try?